The problem of cocaine dependence remains a major medical, social, and legal concern, with 2.3 million Americans estimated to be current users of cocaine. The outcome from medications development efforts has generally been discouraging, excepting for results recently obtained for disulfiram. A series of clinical trials have shown that disulfiram treatment was associated with reduced cocaine use, independent of co-morbid drug or alcohol use. The validation of a human laboratory model against outcomes observed in clinical trials would be a major step forward in developing even more effective treatments for cocaine dependence. This is a crucial undertaking, since there has not been good accord between animal or human laboratory model findings and results from clinical trials. Disulfiram has a variety of actions, one of which is to inhibit dopamine beta hydroxylase, the enzyme responsible for metabolizing dopamine into norepinephrine. Recently obtained data indicate that disulfiram is particularly effective in patients with a "T" allele of dopamine beta hydroxylase; the C/T genotype is associated with reduced enzyme activity compared to the C/C genotype. We therefore propose to evaluate effects of disulfiram treatment on cocaine self-administration in genetically characterized non-treatment seeking volunteers. Participants will be screened and only those with the DBH C/T genotype will be included. Effects of cocaine self-administration will be assessed using two established human laboratory models, one developed at Columbia University and the other developed at Johns Hopkins. Based on the observed effects of disulfiram treatment in clinical trials, especially effects of disulfiram in participants with the C/T genotype, we anticipate that disulfiram treatment will decrease cocaine self-administration in the laboratory.